video semantic annotation

基本解釋

The semantic retrieval of news video is implemented by the annotation of scene through metadata annotation.
通過(guò)引入元數(shù)據(jù)的標(biāo)注完成對(duì)場(chǎng)景的標(biāo)注，實(shí)現(xiàn)用戶的語(yǔ)義查詢(xún)。
In this paper, we discuss the problem of using semi-supervised learning method to do video semantic annotation.
本文討論了利用半監(jiān)督學(xué)習(xí)方法進(jìn)行眡頻語(yǔ)義標(biāo)注的問(wèn)題。
At present, most of video semantic annotation methods are based on statistic theory. The methods use supervised learning method to do semantic label.
目前已有的眡頻語(yǔ)義標(biāo)注方法多是基於統(tǒng)計(jì)學(xué)理論，採(cǎi)用全監(jiān)督學(xué)習(xí)方法進(jìn)行語(yǔ)義標(biāo)注工作。